At the recent Ignite conference, Microsoft announced the public preview of the Azure AI Agent Service. This fully managed service helps developers securely build, deploy, and scale high-quality, extensible AI agents without managing the underlying compute and storage resources.
The Azure AI Agent Service sits within Azure AI Foundry, a trusted platform that enables developers to innovate securely and responsibly with AI. The AI Agent is a smart microservice that answers questions, performs actions, or automates workflows by leveraging generative AI models and real-world data.
(Source: Microsoft Tech Community blog post)
Azure AI Agent Service utilizes the same wire protocol as Azure OpenAI Assistants, allowing developers to create and run an agent using either OpenAI SDKs or Azure AI Foundry SDKs with just a few lines of code. For example, when using the Azure AI Foundry SDK to create an AI agent, a developer needs to specify the AI's model, provide instructions on performing tasks, and define the tools it can access to interact with other services.
Azure AI Agent Service integrates seamlessly with multi-agent orchestration frameworks compatible with the Assistants API, including Microsoft's AutoGen and Semantic Kernel.
The Azure AI Agent Service enables rapid agent development by leveraging OpenAI's Assistants API and over 1400 Azure Logic Apps connectors. It allows agents to perform tasks through various external systems such as Microsoft Teams and Excel. It supports Azure Functions for executing serverless code, performing API calls, and advanced data analysis through a secure Code Interpreter for Python code.
In addition, the service allows for standardized tool connections via OpenAPI 3.0 specifications and extends Llama Stack agents with scalable cloud-hosted tools. Agents can be enhanced with real-time data, private SharePoint documents, and structured data in Microsoft Fabric while offering complete data control through Azure AI Search and Blob storage, further enriched with licensed data sources for tailored responses.
Monalisa Wahlin, a principal product manager at Azure OpenAI Service at Microsoft, writes:
Azure AI Agent Service is flexible and use-case agnostic. This represents endless possibilities to automate routine tasks and unlock new opportunities for knowledge work - whether it is personal productivity agents that send emails and schedule meetings, research agents that continuously monitor market trends and automate report creation, sales agents that can research leads and automatically qualify leads, customer service agents that proactively follow-up with personalized messages, or developer agents that can upgrade your code base or evolve a code repository interactively.
In addition, Shane Larson concludes in a blog post on Azure AI Agent Service:
Azure AI Agent Service is like stepping into the future of business automation. It's not quite the robots-doing-everything future that sci-fi promised us, but it's pretty close! Whether you're trying to improve your customer service, streamline your operations, or make your business run more efficiently, this could be your ticket to the future.
Lastly, more Azure AI Agent Service details are available on the documentation landing page.
